Finding Healthy Fast Foods and Restaurants:¶

df_food.sample(10)

title = "<h2 style='text-align:left;'><b><u><font color='purple'>Sample of Sale Dataframe</font></u></b></h2>"
centered_title = title.center(80)  # Adjust the width as needed

display(HTML(centered_title))
display(df_sale.sample(10))

plt.figure(figsize=(5, 3))
mask = np.triu(np.ones_like(df_food.corr(numeric_only = True), dtype=bool))
sns.heatmap(df_food.corr(numeric_only = True),lw=1,cmap="coolwarm",mask=mask)

plt.show()

df_fat = df_food[['cal_fat','total_fat','sat_fat','trans_fat','cholesterol']]

mask = np.triu(np.ones_like(df_fat.corr(), dtype=bool))

sns.heatmap(df_fat.corr(numeric_only = True),lw=5,cmap="coolwarm",annot=True,mask=mask)

plt.show()

df_food1 = df_food[df_food['item'].str.contains('[5-9]\s*[P p]iece|[1-9]\d+\s*[P p]iece') == False]
print("There are",str(len(df_food1.index)),"foods for one person in the dataframe")

df_food1.sample(10)

average_sodium = df_food1['sodium'].mean()
ax = df_food1.groupby(['restaurant'])['sodium'].mean().sort_values(ascending=False).plot(kind='bar')
ax.set_ylabel('Sodium')
ax.set_xlabel('Restaurant')
ax.bar_label(ax.containers[0], label_type='edge')
ax.set_title('Average Sodium of Meals')

# Add a horizontal line for the average sodium
ax.axhline(y=average_sodium, color='red', linestyle='--', label='Average Sodium')

# Add the value of average sodium on the vertical axis
ax.text(ax.get_xlim()[1] + 0.02, average_sodium, f'{average_sodium:.1f}', ha='left', va='center', color='red', fontweight='bold')# Show the legend
ax.legend()

plt.show()

# Create the pivot table
pivot_table = df_food1.pivot_table(values=['cholesterol', 'sodium', 'calories', 'sugar'],
                                   index='restaurant',
                                   aggfunc={'cholesterol': 'mean', 'sodium': 'median', 'calories': 'mean', 'sugar': 'mean'})

pivot_table = pivot_table.rename(columns={'calories': 'Average<br>Calories', 'sodium': 'Average<br>Sodium', 'cholesterol': 'Average<br>Cholesterol', 'sugar': 'Average<br>Sugar'})

# Define a function to apply the desired color formatting
def color_formatting(val, col_avg):
    if val > col_avg:
        return 'color: red'
    else:
        return 'color: blue'

# Formatting and decoration
formatted_pivot_table = pivot_table.style \
    .format({'Average<br>Calories': '{:.1f}', 'Average<br>Sodium': '{:.1f}', 'Average<br>Cholesterol': '{:.1f}', 'Average<br>Sugar': '{:.1f}'}) \
    .apply(lambda col: [color_formatting(val, col.mean()) for val in col], subset=pd.IndexSlice[:, ['Average<br>Calories', 'Average<br>Sodium', 'Average<br>Cholesterol', 'Average<br>Sugar']])

formatted_pivot_table = formatted_pivot_table.set_properties(**{'font-size': '10pt', 'text-align': 'center'}) \
    .set_table_styles([{'selector': 'th', 'props': [('font-size', '10pt')]}]) \
    .set_caption('Pivot Table - The Average Compounds of Meals')

# Add the note
formatted_pivot_table = formatted_pivot_table.set_caption('Pivot Table - The Average Compounds of Meals<br><br>Note: Red values are greater than the average of each column and blue ones are smaller')

# Display the formatted pivot table
formatted_pivot_table

fig = px.scatter(df_food1, 'cholesterol', 'calories',  color='restaurant', size='sodium', hover_data='item')

# calculate quartiles and interquartile range
q1_x, q3_x = df_food1['cholesterol'].quantile([0.1, 0.9])
iqr_x = q3_x - q1_x
q1_y, q3_y = df_food1['calories'].quantile([0.25, 0.75])
iqr_y = q3_y - q1_y

# identify outliers
x_outliers = (df_food1['cholesterol'] < q1_x - 1.5*iqr_x) | (df_food1['cholesterol'] > q3_x + 1.5*iqr_x)
y_outliers =    (df_food1['calories'] < q1_y - 1.5*iqr_y) | (df_food1['calories']    > q3_y + 1.5*iqr_y)
xy_outliers = df_food1[x_outliers & y_outliers]

# add annotations for outliers
for index, row in xy_outliers.iterrows():
    fig.add_annotation(
        x=row['cholesterol'], y=row['calories'], text=row['item'],
        showarrow=True, arrowhead=1, ax=-50, ay=-40
    )

fig.show()

fig = px.scatter(df_food1[(df_food1['cholesterol'] < 200) & (df_food1['calories'] < 1400)], 'cholesterol', 'calories',  color='restaurant', size='sodium', hover_data='item')
fig.show()

# https://kanoki.org/2020/01/21/pandas-dataframe-filter-with-multiple-conditions/
filtered = df_food1.query('calories > 1000 & cholesterol<300 & sodium<2300')
len(filtered. index)

# df_unhealthy
 
print(str(len(filtered. index)) + " Meals out of " + str(len(df_food1. index)) + " Meals are Considered Healthy" + " ("+ str(round(len(filtered. index)/len(df_food1. index),2)*100)+"%"+")")
# darsad

fig = px.scatter(filtered, 'sodium', 'calories',  color='restaurant', size='cholesterol', hover_data='item')

for i, row in filtered.iterrows():
    
    if row['item']=='Super Sonic Double Cheeseburger W/ Mustard':
        fig.add_annotation(x=row['sodium'], y=row['calories'], text=row['item'], showarrow=True,
                       arrowhead=2, ax=95, ay=95, font=dict(size=8))
    elif row['item']=='Super Sonic Jalapeno Double Cheeseburger':
        fig.add_annotation(x=row['sodium'], y=row['calories'], text=row['item'], showarrow=True,
                       arrowhead=2, ax=40, ay=-40, font=dict(size=8))
    else:
        fig.add_annotation(x=row['sodium'], y=row['calories'], text=row['item'], showarrow=True,
                       arrowhead=2, ax=45, ay=45, font=dict(size=8))

fig.show()

df_salad = df_food[df_food['item'].str.contains("salad", flags=re.IGNORECASE) == True]
print("There are",len(df_salad. index),"salads in the dataset")
df_salad.sample(10)

fig = px.scatter(df_salad, 'calories', 'cholesterol',  color='restaurant', size='sodium', hover_data='item')
for i, row in df_salad.iterrows():
    
    if row['cholesterol']==150:
        fig.add_annotation(x=row['calories'], y=row['cholesterol'], text=row['item'], showarrow=True,
                       arrowhead=2, ax=-40, ay=40, font=dict(size=10))
        
    elif row['calories']==840:
        fig.add_annotation(x=row['calories'], y=row['cholesterol'], text=row['item'], showarrow=True,
                       arrowhead=2, ax=-20, ay=80, font=dict(size=10))
fig.show()

fig, axs = plt.subplots(ncols=3, figsize=(12, 4), gridspec_kw={'width_ratios': [2, 2, 3]})

sns.boxplot(y='restaurant', x='cholesterol', data=df_food, ax=axs[0], showfliers=False)
sns.boxplot(y='restaurant', x='sodium', data=df_food, ax=axs[1], showfliers=False)
sns.boxplot(y='restaurant', x='sugar', data=df_food, ax=axs[2], showfliers=False)

axs[0].xaxis.set_major_locator(FixedLocator(axs[0].get_xticks()))
axs[0].tick_params(axis='y', labelleft=True)

axs[1].xaxis.set_major_locator(FixedLocator(axs[1].get_xticks()))
axs[1].tick_params(axis='y', labelleft=False)

axs[2].xaxis.set_major_locator(FixedLocator(axs[2].get_xticks()))
axs[2].tick_params(axis='y', labelleft=False)

axs[1].set_ylabel('')  # Hide y-axis label for the second plot
axs[2].set_ylabel('')  # Hide y-axis label for the third plot

axs[0].set_title('')  # Remove title for the first plot
axs[1].set_title('')  # Remove title for the second plot
axs[2].set_title('')  # Remove title for the third plot

plt.subplots_adjust(wspace=0.3)  # Adjust the spacing between subplots

plt.show()

df_sale_sorted = df_sale.sort_values('average_sales', ascending=True)
num_bars = len(df_sale_sorted)
fig, ax = plt.subplots(figsize=(16, 9))
ax.barh(df_sale_sorted['restaurant'], df_sale_sorted['average_sales'])
plt.xlabel('Annual Income')
plt.ylabel('Restaurants')
plt.title('Average Sales of the Restaurants')
for i, v in enumerate(df_sale_sorted['average_sales']):
    ax.text(v, i, '${:.0f}'.format(v), ha='left', va='center')
plt.show()

df_veggie = df_food1[df_food1['item'].str.contains('Veg',flags=re.IGNORECASE) == True]
print("There are",str(len(df_veggie.index)),"veggie foods in the dataframe")

title = "<h2 style='text-align:center;'><b><u><font color='purple'>Sample of Veggie Items (13 rows)</font></u></b></h2>"
centered_title = title.center(80)  # Adjust the width as needed

display(HTML(centered_title))
display(df_veggie.sample(13))

df_regular = df_food1[df_food1['item'].str.contains('Veg',flags=re.IGNORECASE) == False]
print("There are",str(len(df_regular.index)),"regular foods (for one person) in the dataframe")

fig, axs = plt.subplots(ncols=1, nrows=3, figsize=(8, 6), sharey=True)

sns.boxplot(y='restaurant', x='cholesterol', data=df_veggie, ax=axs[0], showfliers=False)
sns.boxplot(y='restaurant', x='sodium', data=df_veggie, ax=axs[1], showfliers=False)
sns.boxplot(y='restaurant', x='sugar', data=df_veggie, ax=axs[2], showfliers=False)

axs[0].xaxis.set_major_locator(FixedLocator(axs[0].get_xticks()))
axs[0].tick_params(axis='y', labelleft=True)

axs[1].xaxis.set_major_locator(FixedLocator(axs[1].get_xticks()))
axs[1].tick_params(axis='y', labelleft=True)

axs[2].xaxis.set_major_locator(FixedLocator(axs[2].get_xticks()))
axs[2].tick_params(axis='y', labelleft=True)

axs[0].set_ylabel('Restaurants')
axs[1].set_ylabel('Restaurants')
axs[2].set_ylabel('Restaurants')

plt.subplots_adjust(hspace=0.6)  # Adjust the spacing between subplots

plt.show()

import plotly.subplots as sp
import plotly.graph_objects as go
import pandas as pd

# Create subplots with 1 row and 2 columns
fig = sp.make_subplots(rows=1, cols=2, subplot_titles=("Ordinary", "Giant"), horizontal_spacing=0.1,
                       column_widths=[0.45, 0.45], shared_yaxes=False,
                       specs=[[{'type': 'scatter'}, {'type': 'scatter'}]])

# Add the first scatter plot to the first subplot and customize hover information
df_sale_ord=df_sale[df_sale['us_sales'] < 5000]
fig.add_trace(go.Scatter(x=df_sale_ord['unit_count'], y=df_sale_ord['us_sales'], mode='markers',
                         marker=dict(size=15, color=df_sale_ord['average_sales'], showscale=True,
                                     colorbar=dict(x=-0.25, y=0.5, title='Average Sales')),
                         hovertemplate='%{text}<br>Average Sale: %{customdata}',
                         text=df_sale_ord['restaurant'],
                         customdata=df_sale_ord['average_sales'],
                         showlegend=False),
              row=1, col=1)

# Add the second scatter plot to the second subplot and customize hover information
df_sale_gia=df_sale[df_sale['us_sales'] >= 5000]
fig.add_trace(go.Scatter(x=df_sale_gia['unit_count'], y=df_sale_gia['us_sales'], mode='markers',
                         marker=dict(size=15, color=df_sale_gia['average_sales'], showscale=True,
                                     colorbar=dict(x=1.05, y=0.5, title='Average Sales')),
                         hovertemplate='%{text}<br>Average Sale: %{customdata}',
                         text=df_sale_gia['restaurant'],
                         customdata=df_sale_gia['average_sales'],
                         showlegend=False),
              row=1, col=2)

# Update the layout of the subplots with axis titles
fig.update_layout(height=500, width=900)

# Add y-axis titles to subplots
fig.update_yaxes(title_text="US Sales", row=1, col=1)

# Add x-axis titles to subplots
fig.update_xaxes(title_text="Unit Count", row=1, col=1)
fig.update_xaxes(title_text="Unit Count", row=1, col=2)

# Show the combined scatter plots
fig.show()